Skip to main content

Integrated Information does not equate to consciousness on its own

A recent article:

Makes the claim that the Integrated Information Theory of  consciousness makes it possible that the United States a nation of millions of people is also conscious.

What a strange correlation. I've read Tononi's IIT paper and though it is a brilliant model  for attempting to describe the minutia of consciousness and experience in an agent gathering events via samples from a multi dimensional sensory space...I don't see any reason to suppose that this means the model would apply to all *physical* devices that could through rough mathematical analysis of connections between independent agents also be modeled by an "experience" theory. In fact the author of the article above actually references this caveat as made by Tonini in his paper but then goes to interpret it incorrectly:

"Before we saddle Tononi straightaway with commitment to the consciousness of the United States, though, there is one issue to address: Despite the liberality of his view, Tononi does not regard every putative system as an “entity” that could be the locus of consciousness. If a putative system contains no causal, that is, informational, connections between its parts, then it is not an entity in the relevant sense; it is not, he says, a “complex." Also, a putative system is not a conscious entity or complex if a larger, more informationally integrated system entirely subsumes it. For example, two disparate nodes do not constitute a conscious complex if a third node lies between them creating a more informationally integrated network. This restriction on the possible loci of consciousness is still extremely liberal by commonsense standards: Complexes can nest and overlap, for example, within the brain, where tightly integrated subsystems interact within larger less-integrated systems."

The author is making a critical assumption with is already false, he's assuming that the critical requirements of specific cascades between memory storage elements (neurons and glia in the brain as analog to humans proper in the country) and specific connections between suites of memory storage elements (the circuitry of connection between regions that enable analysis, comparison and autonomic and emotional modulation of *intention) does not exist in precise and necessary ways.

We already know from the many fMRI studies coming out that it does! If you don't connect the regions in the *right* way, whatever that won't get conscious emergence. Over lapping of some subnetworks means nothing if you don't get the *right* over lapping. A most recent study was able to actually show the precise way that different regions within the brain became connected or disconnected at the dawn of consciousness itself! These are huge huge clues to those investigating consciousness in order to at least figure out the correct ways to *connect* different regions of processing in an effort emerge it in our attempts to build synthetic dynamic cognitive agents.

The United States is a randomly connected set of networks...most of which absolutely greedy in their connection patterns. Most of which oblivious in how they "fire" such that cascades tend to persist only across local networks of a way that is very different from the brain...which was *evolved* to enable efficient cascades from stimulus to analysis to comparison to import.

It stuns me that some one would make the case that the author is doing in this piece. I believe he is totally misinterpreting Tonini's statements regarding comparing "consciousness" between living agents and non living agents (like a diode).

Only some one making a really tight read would correlate possible states of input with "consciousness". Consciousness is NOT just about dimensional variation in a state of input (in the biological analog a dimension of sensation) and I never in my read of his views or viewing his talks thought he was giving that view. Tonini is presenting the abstract general representational model (integrated information...which I upon reflection is trivially true) but does NOT define the precise and necessary underlying constructed circuitry for creating that model as a dynamic process as it IS in the minds of conscious beings.

A diode will just sit there absent input but our neurons are constantly in motion and in action because of the dance (a precise dance) of enabled circuit connections between different regions that I posit requires the music of intention (autonomics + emotion) , memory, incoming sensory input across multiple dimensions of sensation and comparison of those sensations to memory as modulated by the intentional tuple.

I would give that perhaps some of the blame for the interpretation of Tonini's theory as correlating consciousness of the mammalian kind to simple information integration is due to Tonini's some what loose use of the word for both static and dynamic connected systems with precise and necessary connections between dynamic subsystems for emerging selective cognition. One absent inputs just sits there, the other roils in its own internal dialogue absent of inputs. I'd posit a great way to know if we've created consciousness is to look for dreaming when the agent is not actively taking sensory input...dreams persist the internal dialogue of dynamic cognition and I assert can only be had by minds with the necessary and precise internal connections for them. I've provided some ideas on dreams (the biological function supposed is as repair process)  in another blog post.

Last year I diagrammed a model that I believe will be successful in emerging consciousness of a biological type should all the necessary connections and algorithms of modulation be found. I hope to start work actually constructing the first such example of what I call a synthetic dynamic cognitive agent in the next couple of years.



Matt Sigl said…
Just a quick note on your thoughts. I can't say I followed your entire blog post (more due to confusion on my part than with your writing, I'm sure.) but Tononi is quite clear that his theory does apply to any physical object. To be more specific, he makes it clear that where there is causation, there is information, and where there is information, there is consciousness. Since there is causation everywhere (and at every physical scale from quantum to neuronal to universal) there is consciousness everywhere, hence panpsychism. I'm not extrapolating; Tononi is quite clear about the theory's panpsychist implications, as is Christof Koch, another IIT defender. How different is a simple experience (say of 1 bit) to our daily consciousness? Well, how different is our whole body to single atom? That's probably a fair comparison.

Information integrates at different timescales, in neurons it looks like it's somewhere around 300-700 milliseconds, and so humans have a consciousness that runs at this rate. It is a prediction of the theory that any collection of elements that have integrated causal interactions over a certain time-scale will, at some level, generate conscious experience. The conscious experience's character is defined wholly by the abstract informational relationships generated by the system. In fact what the system IS (neurons, transistors, planets) is irrelevant to the experience just as Microsoft Word doesn't care which brand of PC it is on.

That being said, it seems given the random "connectome" of most collections of non-neural entities, the phi value (the measure of the amount of consciousness generated) is either zero or infinitesimally small. It took eons of evolution to carve out of nature an information processor that could generate such complicated information structures. The people of the United States as a whole may indeed generate a conscious experience, but that experience would most likely be mind-numbingly basic. More interesting is the way the quantum world integrates information both with the wave-function collapse and with the concept of entanglement. Tononi briefly and tantalizingly addresses these issues in a footnote in his manifesto.

You bring up memory. Although memory is not necessary for consciousness according to the IIT, experiments done on computer systems in an attempt to generate phi show that applying memory functions to the system as it programs itself can help it create causal structures that DO amplify phi growth. Memory appears to be an important aspect of how nature created consciousness even if it is not, strictly speaking, necessary for any particular instance of consciousness.

As far as neural circuity goes, the theory makes a specific prediction. It claims that in any conscious human experience there should be a network of neurons in the brain connected in such a way that, when activated, engages a causal structure that cannot be reduced to the sum of its parts. While uncovering this "central" neuronal object is a challenge, in theory it should nonetheless be empirically verifiable.

Anyway, glad you wrote the post. Science is finally getting around to the mind and it's great to see people paying attention.
David said…
Great points Matt, regarding memory I see it as important for our type of consciousness which is a dynamic consciousness. Rather than being snapshots of moments taken in response loosely to stimuli (as say it surely is for a gnat or dragon fly) ours is a running *internal* and external dialogue and that necessarily factors into the network of neuronal connections being re-weighted across the levels of cognition involved.

To me we see our model quite clearly in the various structures of the brain as they differ between different animals and an appeal to comparative brain science can help us quickly separate the wheat from the chaff as it were. At least in trying to understand what I call "emotional resolution" a key factor that we'll need to master if we are to create stable cognitive rather than pathological agents.

I wrote a bit more on the natural connection in this blog post from february:

Popular posts from this blog

On the idea of "world wide mush" resulting from "open" development models

A recent article posted in the Wall Street Journal posits that the collectivization of various types of goods or services created by the internet is long term a damaging trend for human societies.

I think that the author misses truths that have been in place that show that collectivization is not a process that started with the internet but has been with us since we started inventing things.

It seems that Mr. Lanier is not properly defining the contexts under which different problems can benefit or suffer from collectivization. He speaks in general terms of the loss of the potential for creators to extract profit from their work but misses that this is and was true of human civilization since we first picked up a rock to use as a crude hammer. New things make old things obsolete and people MUST adapt to what is displaced (be it a former human performance of that task or use of an older product) so as to main…

Engineers versus Programmers

I have found as more non formally trained people enter the coding space, the quality of code that results varies in an interesting way.

The formalities of learning to code in a structured course at University involve often strong focus on "correctness" and efficiency in the form of big O representations for the algorithms created.

Much less focus tends to be placed on what I'll call practical programming, which is the type of code that engineers (note I didn't use "programmers" on purpose) must learn to write.

Programmers are what Universities create, students that can take a defined development environment and within in write an algorithm for computing some sequence or traversing a tree or encoding and decoding a string. Efficiency and invariant rules are guiding development missions. Execution time for creating the solution is often a week or more depending on the professor and their style of teaching code and giving out problems. This type of coding is devo…

1905: Annus Mirabilus - Photo electric effect

1905 was a great year for physics, in this year a 24 year old patent examiner in Bern Switzerland published 4 fundamental papers in physics in 4 disparate areas of the field. The topics included special relativity, the relationship between energy and matter, brownian motion and the subject of this post, the photo electric effect.

The photo electric effect paper by Einstein was probably the most practical paper next to the brownian motion paper in that it provided an answer to a long standing problem in electromagnetic theory at the time that had stood as an embarrassment to particle physics. This embarrasment was a legacy of the work of James Clerk Maxwell and his fundamental equations of electromagnetism, by using a continuous wave analog to describe the energy of propagating fields Maxwell was able to do the astonishing, he explained the riddle that was the relationship between electricity and magnetism in clear mathematical terms and he was able to show how light must be itself an …